Building Blast

7/29/2019 Building Blast

1/6I January 200820

Building Blast Simulation

and Progressive CollapseAnalysis

FEA analysis of severe blast loading supports the design of survivable structures without necessarilyrequiring expensive physical simulations of a specific explosive or combustion event. These analyses

also show that using established structural design guidelines may not be conservative for severe

blast loading on steel structural members.

This article, from T. Krauthammer of University of Florida, and J. Cipolla of Simulia, describes the FEA

modeling of the progressive collapse of a steel frame structure, and the qualitative insights gained. A full-

length version of this paper is available in the NAFEMS 2007 World Congress proceedings.

T. Krauthammer, University of Florida and J. Cipolla, SIMULIA, Inc.

Analysis of steel frame

connections under blast loadsCurrently, U.S. design guidelines for steel

connections in structures subjected to blast loadsare based on recommendations in the Department

of Defense Technical Manual (TM) 5-1300TM [1].

The approach idealizes real structures and

structural elements as equivalent lumped-mass

single-degree freedom systems. In addition, the

guidelines are for single-storied steel frames, notsubjected to any significant dead loads apart from

the self weight of the structure. The actual effects of

blast and dead loads on real steel connections may

exceed the margins predicted by TM 5-1300.

To assess the behaviours of steel moment

connections under such loads [2], finite-element

simulations using ABAQUS/Explicit 6.5 [4] wereemployed. The maximum rotational capacities of

the connections were then compared against

values derived with the TM 5-1300 approach [1].

Target connections were placed between beam and

column at the ground floor of a multi-story building.

Lengths of the beams and columns were taken

from Engelhardts, et al. [8], experimental studies

(Figure 1). For each connection type, four differentload cases were used (see Table 1).

Reference maximum blast pressures were

calculated based upon the TM 5-1300 criteria [1].

For the finite-element calculations, the standalone

codes SHOCK [9] and FRANG [10] were used to

compute equivalent shock pressures and gas

pressures (Table 1), assuming an 18.5 lb. TNT

charge located at the centre of the room. Weapplied blast pressures as spatially uniform surface

loads on the sidewalls that transferred to the beams

and the column of the connection. Dead loads,

applied on the top flanges of the beams and axially

on the top cross section of the column, correspond

to those for a 10-stor y office building. The

numerical model was analyzed with and withoutthese dead loads to evaluate their influence on

connection response.

An isotropic elasto-plastic model was used as the

material property for each connection component

[2]. Yield and ultimate strengths were increased to

account for strain rate effects using dynamicincrease factors (DIF) as recommended in TM 5-

1300 [1]. Since brittle fracture on the weld

connections was anticipated under the blast loads,

we adopted the shear failure model; ABAQUS

removed elements from the mesh as they failed.

The finite-element models (Figure 2) were created

using predominantly 8-noded continuum brick

elements with reduced integration.

The responses and failure criteria based on TM 5-

1300 criteria are shown in Table 2, and indicate

that the representative room could withstand the

loads from the explosive charge.

the representative room could withstand

the loads from the explosive charge.


2/6January 2008 I 21

787 Final Assembly Factory Flow

Figure 1:

Geometrical Modelused for Numerical

Study

Figure 2: Representative Finite-element Model of Steel Connection

This type of intrinsically

transient nonlinear

phenomenon is difficult

to model, understand, ordesign against without

finite-element analysis.


3/6I January 200822

Finite-element results when blast pressures (Table 1) were

applied to floor and sidewalls are summarized illustrated inTable 3. The predicted global rotations of the beams are

close to the TM 5-1300 results for the frangible wall cases.

However, the beams where the reflecting walls were

located rotated much more than TM 5-1300 computation

predicted, with the result that a greater impulse and energy

are transferred to the beam and column in the room.

Sample maximum local rotations, associated with the

plastic hinge that is formed in the beam, are shown in

Figure 3. All local rotations for the different cases exceeded

the limit of 2 degrees specified in TM 5-1300.

Note, also, that the beams twisted more severely

horizontally, and clearly exceeded the TM 5-1300 limit

criteria, since realistic internal blast pressures radiateoutward in three dimensions. These findings indicate that

severe damage in the connections comes from the blast

radiating in three dimensions as well as the verticallyapplied pressure. Deformation data for beams and column

in the various cases indicate that dead loads and DIFs

enhanced structural strength, but the beam cross sections

twisted additionally due to dead loads. The column

rotations indicate that the columns did not significantly

affect the connection damage. According to the stress and

strain results, components in all connections yielded for allthe cases.

These analyses show the value of investigating structural

connections using high-resolution finite-element analysis.

For example, a steel moment connection judged safe

based on TM 5-1300 criteria failed in the finite-element

simulations. Moreover, TM 5-1300 criteria may need

revision to reflect findings based on more complexbehaviour.

Figure 3: Global vs. Local Rotations Case 1, No DL, No DIF. Horizontal (left) and Vertical (right)


4/6

Figure 5: Case 6 with Ideal Connections

January 2008 I 23

Progressive collapse of steel

frame structuresProgressive collapse is a failure sequence in which local

damage leads to large scale collapse in a structure. It has

been an important issue in building design since the

collapse of the Ronan Point apartment building in 1968

[11]. This type of intrinsically transient nonlinear

phenomenon is difficult to model, understand, or designagainst without finite-element analysis.

Ten-story 3D moment frames with rigid and semi-rigid

connections were studied for their sensitivity to failure of

specific columns [3]. Three failure modes were considered:

material, buckling, and connection failures. The first two

have been studied extensively elsewhere [12, 13, and 14].

Experiments show that a real steel connection is neitherrigid nor pinned [15]. In this study, the nonlinear moment-

rotation relationship of the 10-story frame was obtained

through extensive preliminary 3D finite-element simulation

of steel connections.

Six initial failures with rigid and semi-rigid connections wereused to analyze the frames for progressive collapse of five

stories. Frame columns were based on a simple Load andResistance Factor Design (LRFD) procedure manual [16].

Both ideal (rigid plus hinge) and semi-rigid connections

were adopted for the progressive collapse analyses.

Analyses were performed up to seven seconds after the

initial failure, modeled by instantaneous removal of a

designated column. The failure cases are shown in Figure4. Only Case 6, where three columns were removed,

caused total collapse of the building. Figure 5 shows the

result for Case 6 of the building with ideal connections.

Case 6 with semi-rigid connections also collapsed, but

differently, as shown in Figure 6. The first failure was

initiated at a connection, as shown in Figure 6-(b). As

connections failed, the floors above the removed columnsstarted to fall to the ground, and it caused columns

buckling in the 6th floor, as shown in Figure 6-(c). These

standard design

criteria may not be

conservative

enough

column buckling cases initiated horizontal failure

propagation in the 6th floor, and the whole floor failed.

After that, the columns in the first floor buckled because

the floors collapsed, leading to the total collapse of thebuilding.

The 10-story frame, designed for gravity and lateral loads,

performed fairly well in the simulations. Even though the

ideal and semi-rigid connection cases both caused total

collapse for Case 6, they showed very different qualitative

behaviour. The collapse of the semi-rigid connection casewas caused by a cascade of local failures, such as

connection failures and columns buckling. However, the

collapse of the ideal connection case was caused by

column buckling in the first floor. These different failure

mechanisms are quite apparent in the nonlinear finite-

element results.

The analyses also showed that once failure propagationinitiated (i.e. horizontal column buckling), it would not stop

until it caused total, or almost total, collapse. To protect a

structure against progressive collapse, horizontal column

buckling propagation appears to be the most critical factor

to control.

Figure 4: Initial Column Failure Cases


5/6I January 200824

ConclusionsFEA analyses reliably simulate critical aspects of

structural behaviour, such as the details of steel

connection designs and failure modes.

Our analyses suggest that connection behaviour under

blast loading can vary significantly from standard design

criteria. Importantly, the simulations also suggest thatstandard design criteria may not be conservative enough

for the cases modeled here and may require refinement

and revision in light of nonlinear transient effects, suchas progressive collapse. Finite-element analysis of

progressive collapse due to blast effects also reveals

qualitative information about structural failure such as,

in these cases, sensitivity of failure mode to connections.

ContactJeffrey Cipolla

[email protected]

To protect a structure against

progressive collapse, horizontal

column buckling propagation

appears to be the most critical

factor to control.

Figure 6: Case 6 with Semi-rigid Connections

REFERENCES[1] Structures to Resist the Effects of Accidental Explosions TM 5-1300,

Department of the Army (U.S.), 1990.

[2] H. C YIM, C. STARR, T. KRAUTHAMMER, AND J. LIM. Assessment

of Steel Moment Connections for Blast Loads, Proc. 2nd InternationalConference on Design and Analysis of Protective Structures 2006,

Singapore, 13-15 November 2006.

[3] T. KRAUTHAMMER, H.C. YIM, C. STARR, AND J.H. LIM. Progressive

Collapse of Multi-Story Steel Frame Structures, Proc. 32nd DoDExplosive Safety Seminar, Philadelphia, PA, 22-24 August 2006.

[4] ABAQUS Analysis Users Manual, Version 6.5, ABAQUS, Inc.,ABAQUS, 2005.

[5] Fundamentals of protective design for conventional weapons TM 5-

855-1, Department of the Army (U.S.), 1992.

[6] RIPLEY, R.C., DONAHUE, L., ZHANG, F. Modelling Complex Blast

Loading in Streets, 6th Asia-Pacific Conference on Shock & Impact

Loads on Structures, Perth, Australia, December 7 - 9, 2005.

[7] J. CIPOLLA, Generalized incident wave loading on acoustic and solidelements, Proc. 77th Shock and Vibration Symposium, 2006.

[8] ENGELHARDT, M.D., SABOL, T.A., ABOUTAHA R.S., AND FRANK

K.H., Northridge Moment Connection Test Program Report for AISC,

1994.

[9] NCEL, SHOCK Users Manual, Naval Civil Engineering Lab., Port

Hueneme, CA, 1988.

[10] WAGER, P., AND CONNETT, J., FRANG Users Manual, Naval Civil

Engineering Lab., Port Hueneme, CA., 1989.

[11] GRIFFITHS, H., PUGSLEY, A., SAUNDERS, O., Collapse of Flats atRonan Point, Canning Town, Her Majesty s Stationery Office London,

1968.

[12] ARISTIZABAL-OCHOA, J. D., Elastic Stability of Beam-Columns with

Flexural Connections under Various Conservative End Axial Forces,Journal of Structural Engineering, Vol. 123, No. 9, pp. 1194-1200,

September 1997.

[13] ERMOPOULOS, J. CH., Buckling Length of Framed Compression

Members with Semi-rigid Connections, Journal of Constructional

Steel Research, Vol. 18, pp. 139-154, 1991.

[14] LIEW, J. Y. R., CHEN, W. F., CHEN, H. Advanced Inelastic Analysis ofFrame Structures, Journal of Constructional Steel Research 55, pp.

245-265, 2000.

[15] KAMESHKI, E. S., SAKA, M. P., Genetic Algorithm Based Optimum

Design of Nonlinear Planar Steel Frames with Various Semi-Rigid

Connections, Journal of Constructional Steel Research 59, pp. 109-134, 2003.

[16] AISC, Load and Resistance Factor Design Structural Members,

Specifications, & Codes, American Institute of Steel Construction,

1994.


6/6

Nasmyth Building Scottish Enterprise Technology Park East Kilbride G75 0QR Scotland, UK

T +44 (0)13 55 22 56 88 F +44 (0)13 55 24 91 42 E [email protected] Wwww.nafems.org

Creating Awareness I Delivering Education I Stimulating Standards

NAFEMS is the International Association for the

Engineering Analysis Community, an independent,

not-for-profit, international membership association,owned by its members. The scope of its activities

encompasses all simulation technology, including

Finite Element Analysis and Computational Fluid

Dynamics. As new application areas and techniques

constantly evolve, NAFEMS becomes involved to

create awareness and deliver appropriate education

and training.

NAFEMS publications and benchmarks are widely

regarded within the engineering analysis community

as the most authoritative source of information

available. The areas covered by NAFEMS are

expanding year by year with the growth in

membership, and people increasingly view NAFEMSas a one-stop shop for all aspects of information on

engineering analysis.

For engineering analysts, NAFEMS offers an

excellent platform for continuous professional

development.

Get Involved.Join NAFEMS Today.

NAFEMS Membership is a

must for all those truly

involved with numerical

analysis, for continuous

improvement and learning

and sharing of experience

Instituto Technologico de Aragon

Documents

Building Blast